The long term objectives of this project are to understand further how people know and come to know what factors influence their happiness. Past research shows that when thinking about how a future event will influence happiness, people often overestimate the duration of their affective responses (called the durability bias). The specific aim of the present proposal is to develop a comprehensive model that predicts when people will use information about their past reactions to make more accurate forecasts. The model tested here posits that information about the past must be accurate (i.e., available), brought to mind (i.e., accessible) and applied (i.e., applicable). The reason the durability bias exists is that these conditions are rarely met. To test the validity of this model, controlled experiments are proposed which will test the effect of thinking about past experience and the applicability of past experience on affective forecasts. A field study will assess some of the everyday implications of the model. The primary benefit of this research is the enhancement of the field's knowledge about intuitive hedonics, or how well people know what influences their emotions. Affective forecasts implicitly or explicitly influence behavior and decisions that are important to people (e.g., decisions about career choice). Thus, understanding these processes has implications in such fields of study as judgment and decision making and health psychology.